A Viable and Scalable Route for the Homogrowth of Si Nanoconesand Si/C NanostructuresS. Orlanducci, F. Toschi, V. Guglielmotti, I. Cianchetta, C. Magni, E. Tamburri, and M. L. Terranova

Department of Science and Chemical Technologies and MINIMA Lab, University of Rome “Tor Vergata”, Via della RicercaScientifica, 00133, Rome, Italy

R. Matassa‡ and M. Rossi*,‡,§

‡Department of Basic and Applied Sciences for Engineering, Sapienza University of Rome, Via A. Scarpa 14, 00161 Rome, Italy§Research Center of Nanotechnologies for Engineering (CNIS), Sapienza University of Rome, P.le A. Moro 5, 00185 Rome, Italy

ABSTRACT: The present paper focuses on the fabrication ofpolycrystalline Si nanocones, starting from hydrogen-containinggas mixtures activated in a dual-mode microwave/radio frequencyplasma reactor. Taking advantage of the capacitive coupling ofmicrowave (MW) and radio frequency (RF) plasmas, densearrays of vertically aligned or isolated nanocones, withpredetermined basal diameters, can be routinely prepared byengineering the distribution of Ni sites on lithographicallypatterned Si surfaces. Such elongated objects are constituted bynanocrystalline Si extruded from the negatively biased substrate. Under suitable experimental conditions, a thin surface deposit ofhydrogenated amorphous C is achieved. The one-step growth of elongated three-dimensional Si/C entities is rationalizedconsidering the Si−Ni diffusion induced by energetic H ions and the formation of Si/Ni systems with a melting temperaturelower than the bulk Si. Reflection high-energy electron diffraction (RHEED) analysis and micro-Raman spectroscopy have beenused to study the crystallographic features of the Si nanocones and to confirm the role played by the Ni deposits. The presentmethodology is proposed as a viable route for a scalable fabrication of designed Si or Si/C nanostructures to be integrated in Sidevices.

■ INTRODUCTION

Over the last two decades, a strong downscaling of electronicshas taken place, supported and funded especially by theinformation technology industry, interested to expand themainstream Si-based technology by developing new, high-performance devices. This continuous scaling is driving towardthe realization of novel device architectures based on theassemblies of different nanosized Si structures. Among a varietyof proposed Si nanostructures, nanowires and nanocones arepresently considered as the most attractive components notonly for micro- nanoelectronics1,2 and optoelectronics3 but alsofor a variety of other applications, ranging from batteries4 tosolar cells5 and cold cathodes.6 Structures with a high aspectratio, characterized by size confinement in two coordinates,offer indeed a number of advantages with respect toconventional bulk or thin film counterparts, such as size andshape control, lower power consumes, and higher integration indevices.However, the fabrication of Si-based ever-small electronic

and opto-electronic devices could cease to be worthwhile, firstbecause it cannot be guaranteed that Si at much smaller lengthscale will behave as expected, and also because of the costs forfabrication of complex structures using top-down lithographicprocesses. Nanoscale Si is a complex material with regard to its

stability, because the optoelectronic properties show environ-mental and aging degradation, with resulting ambiguity indevice performance evaluation.Alternatives to all-Si nanoelectronics will involve develop-

ment of hybrid devices, a combination of foreign nanostruc-tures atop Si platforms. The most suitable candidate forfabrication of heterostructures is without any doubt theversatile C, another element of the IV group of the PeriodicTable that can be produced in a variety of forms, fromamorphous films to highly crystalline structures, both at themacro-, meso-, and nanoscale. Specific carbon forms can addnovel functionalities, exploiting capabilities of Si-based devices.In this context it is thought that the coupling of Si

nanostructures with carbon-based nanomaterials could help inovercoming, at least partially, some critical limitations. This is avery attractive prospect, but it is not an easy task and the wayforward is to systematically study processing methods aimed tobuild up different combinations of Si and C.Tailored Si and C complex nanostructures are nowadays

designed and obtained mainly using chemical vapor deposition

Received: May 8, 2012Revised: June 18, 2012Published: July 25, 2012

Article

pubs.acs.org/crystal

© 2012 American Chemical Society 4473 dx.doi.org/10.1021/cg3006212 | Cryst. Growth Des. 2012, 12, 4473−4478

(CVD)-based apparatuses, where kinetically driven processespromote the creation of novel materials, enabling at the sametime to modulate their size, shape, and structure.However, processes carried out under nonequilibrium

conditions often reveal surprises, due to the dynamiccompetition between concomitant reaction routes. Theexperimental evidence suggests that many aspects of alreadysettled synthesis methodologies are still poorly understood andthat some less considered phenomena could be just asimportant in determining the nature and nanostructure of thefinal products as the main deposition conditions usually takeninto consideration.Using a dual mode microwave−radio frequency (MW-RF)

reactor in order to implement a bottom-up nanofabricationmethod, a systematic study is currently going on in ourlaboratories with the general target of building up three-dimensional (3D) Si and hybrid Si/C nanostructures. Inparticular, the growth of conical-shaped and vertically aligned Sinanostructures with amorphous hydrogenated carbon surfacelayers is here reported. Such specific hybrid nanostructurescould be useful in a variety of applications, including therealization of high-capacity anodes for Li-ion batteries, havingthe potential capability to increase electronic contact andconduction and buffer volume changes during Li-ion insertion/extraction.7

■ EXPERIMENTAL SECTIONThe Si/C structures were generated using a purpose-designed plasmaenhanced-chemical vapor deposition (PE-CVD) reactor, where the gasphase (a H2/CH4 mixture) is excited by a dual-mode MW/RF plasma.A detailed description of the apparatus can be found in ref 8. In thepresent experiments two frequencies, MW (2.45 GHz) and RF (13.56MHz), were applied to initialize the plasma.This technique connects the advantages of microwave discharge,

such as homogeneity of plasma and high concentration of activespecies, with the ability to control ion energy and directionality byproper biasing of the substrate.9 In particular, it is well-known10 the RFinduces a negative DC self-bias potential Vb (measured with respect tothe ground) that forces the accelerated positive ions to travel towardthe substrate following the lines of the applied field. The Vb values aredependent on both the RF power and on the working pressure. In the5−20 mbar range, at an RF power of 100 W, an experimental relationof the type Vb = A + Be−CP was found to hold, with A = 39 ± 1 V, B =529 ± 1 V, and C = 0.100 ± 0.001 mbar−1. In the present experiment,aiming to reach a Vb value of 230 V, the total pressure was kept at 10mbar and the MW and RF powers were fixed at 100 W. On the basisof several trials it was found that, in order to produce Si nanoconeswith a thin C coverage, a suitable ratio between the flowing rates of H2and CH4 could be 10. The flow rates of H2 and CH4 were thereforefixed at 100 and 10 sccm, respectively. The runs lasted 30 min. Duringthe runs the substrate temperature, measured using a thermocoupleattached to the substrate holder, was found to vary from T = 510 ± 10°C at the beginning of the process up to T = 620 ± 10 °C at the end.The synthesis experiments have been performed on Si (100)

substrates patterned by lithography using masks shaped as “4” and ascircular spots of 150 nm diameter, respectively. The substrates wereprepared by sputtering 15 nm thick Ni inside the patterned areas.The deposits have been characterized using scanning electron

microscopy (SEM), reflection high energy electron diffraction(RHEED), transmission electron diffraction (TED), and micro-Raman spectroscopy.SEM observations have been carried out by a field emission SEM

Hitachi S4000. RHEED observations have been performed at 60 keVon an electron optics column AEI EM6G, equipped with a high-resolution diffraction stage.The micro-Raman spectra were acquired using an Ar ion laser

(514.5 nm excitation wavelength) and a 600 g/mm and a 1800 g/mm

gratings spectrometer (iHR550 - HORIBA JOBIN YVON) coupledwith a liquid-nitrogen cooled CCD. The spectral resolution is 3 cm−1

(using the 600 g/mm) and lower than 1 cm−1 (using the 1800 g/mm).In order to avoid samples damage, the laser power was settled at about5 mW.

■ RESULTS AND DISCUSSION

Figure 1a shows the FE-SEM image of a typical deposit growninside a Ni-catalyzed area shaped as “4”. A larger magnificationof the boundary between coated an uncoated regions allowsone to note that the deposits are formed by regular arrays ofvertically elongated structures (Figure 1b,c). The conical-likewell aligned objects (length up to 1.5 μm) are found toprotrude out of the surface at 90°.Taking a series of Raman spectra from various regions of the

vertical deposits, a preliminary structural study has beenperformed. No significant differences were found among suchspectra. A typical as detected and background subtractedRaman spectrum in the range 300−2700 cm−1 of the sample isdisplayed in Figure 1d.The broad signals at about 1350 and 1598 cm−1,

corresponding to the D and G bands of nanographiticstructures, indicate the presence of an amorphous sp2-carbonphase. The strong luminescence noted in all the spectra is likelydue to highly hydrogenated amorphous carbon, produced underour experimental conditions that induce negative self-biasing ofthe substrate.11,12 Under the action of such high negativepotential the hydrogen ions of the plasma are indeed forcedtoward the substrate, where they impact with the growingspecies inducing an effective hydrogenation of the amorphouscarbon.In an attempt to remove the hydrogenated amorphous

carbon phase, the samples were submitted to a thermaltreatment, carried out at 980 ± 20 °C for 2 h in a N2atmosphere. FE-SEM images of the deposits submitted to thethermal treatment are shown in Figure 2a,b. Heating did notsubstantially modify the original architecture of the deposits,and the vertically protruding conical structures survivedmaintaining their shape in terms of height and basal diameter.A partial bending of some of the conical apexes is also observed.From inspection of the Raman spectrum reported in Figure 2c,the complete disappearance in the Raman spectrum of the Dand G bands as well as of the luminescence due to thehydrogenated sp2-carbon can be appreciated.The Raman spectra taken after the thermal treatments show,

for all the samples, only well-defined peaks at about 520 and980 cm−1, related to first- and second-order scattering fromcrystalline Si.13 The comparison of the spectra registered beforeand after the sample heating demonstrates that the conicalobjects generated by the CVD runs consist of crystalline Sinanostructures, having a thin surface layer of amorphoushydrogenated sp2-carbon, completely removed by the heating.It is to be noted that among the scientific community the

interpretation of one phonon Raman spectrum of the Sinanostructures is still under debate. Several reasons have beeninvoked in order to explain the inconsistencies of the Ramanparameters measured for the various Si nanostructures. Suchinconsistencies can be rationalized taking into account sampleinhomogeneity, Fano scattering from photogenerated carriers,stresses and laser induced heating.14−18 For a deeper discussionabout this matter, more detailed Raman spectra have beentaken in the range 450−590 cm−1, using the 1800 g/mmgrating (Figure 2d).

Crystal Growth & Design Article

dx.doi.org/10.1021/cg3006212 | Cryst. Growth Des. 2012, 12, 4473−44784474

Under our experimental conditions, the Si substrate exhibitsa Gaussian line shape with a maximum at 518 cm−1 and a fwhmof 7 cm−1 (spectrum in the bottom of Figure 2d). The

Figure 1. (a−c) FE-SEM images of sample with a Ni-catalyzed areashaped as “4”: (a) general view; (b, c) higher magnifications from theboundary regions; (d) typical micro-Raman spectrum as acquired andafter background subtraction.

Figure 2. (a, b), FE-SEM images at low (a) and higher (b)magnification of the deposits submitted to the thermal treatment; (c)overview of the background subtracted Raman spectra from the samesample, before and after the thermal treatment, in the spectral wideregion of interest for Si and C (300−2000 cm−1); (d) details of theprevious Raman spectra in the range 450−590 cm−1, compared withthe spectrum achieved from the (100) Si substrate in the same spectralrange.

Crystal Growth & Design Article

dx.doi.org/10.1021/cg3006212 | Cryst. Growth Des. 2012, 12, 4473−44784475

occurrence of a slight red shift and of the broadening of the Siband with respect to the typical features of the Si signal (ω =520 cm−1 and fwhm = 3 cm −1) is consistent with heatingeffects.19 On the other hand the spectra obtained from samplebefore and after thermal treatment (Figure 2d) show a largerdownshift and an asymmetrical enlargement of the first-orderband. These curves can be deconvolved in two componentscentered at about 510 and 517 cm−1, and ascribable to differentSi phases existing in the sample. The principal line centered atabout 510 cm−1 is due to the first-order Raman frequency insilicon nanostructures, the large shift observed with respect tothe bulk Si is likely related to laser heating effects coupled witha poor thermal dissipation. With respect to bulk Si, both sizeeffects and poor thermal contact with the substrate can explainthe high sensitivity of the Si nanowires even to the low increaseof temperature produced by a moderate laser power.20

Finally the peak located at 517 cm−1, with a larger bandwidthand a slightly red shift with respect to the (100) Si, arises fromthe defective surface of the silicon substrate. The decrease ofthe 517 cm−1 bandwidth after the thermal treatment can berelated to a partial recrystallization of the silicon surface and animproving of the thermal contact with the silicon nanostruc-tures.In conclusion, the analysis of Raman data highlight that the

nanocones are constituted mainly of nanocrystalline Si. Theelongated shape of the protruding nanostructures wouldaccount for the poor thermal contact with the substrate.A complementary structural investigation was carried out on

the thermally treated samples using a RHEED apparatusequipped with an high resolution goniometer stage, thatenabled us to vary the incidence angle θ of the electron beamwith respect to the sample surface and to perform measure-ments under different geometrical conditions. The RHEEDanalysis did not show significant and appreciable differences,within the sensitivity limits of such diffraction techniques, in thecrystalline features of the samples before and after thermaltreatments. The only difference is the presence, before thethermal treatment, of a diffused background that is typically dueto the presence of a very thin surface amorphous layer.21

The first measurements were performed in selected areaconditions using an aperture of 50 μm and keeping theboundary between coated an uncoated regions parallel to the e-beam. Probing such a region, the RHEED pattern (Figure 3a)shows the concomitant presence of (100) Si single-crystal(spots) and of polycrystalline randomly oriented Si (rings).Further measurements were carried out using a smaller

aperture (10 μm) and decreasing the incidence angle as muchas possible. Moving the e-beam toward the Ni-coated region onwhich the FE-SEM analysis revealed the presence of conicalstructures (see Figure 2), it was possible to drive the e-beam tointeract prevalently with the outermost part of the sample, thusobtaining a diffraction pattern in transmission conditions(Figure 3b) through the protruding structures.Under such conditions the RHEED pattern consists of

slightly and radially broadened diffraction rings, whose spacingsis indicative of polycrystalline Si. The broadening of the ringscan be straightly attributed to the presence of crystallites with asmall size (in the range of very few tens of nanometers) and/orto the concomitant presence of lattice strain due to surfacerelaxation and nonuniform lattice distorsions. Such distortionscould be ascribed to the surface tension of crystallites, keepingin mind that the probability and intensity of the surface tensionincrease with crystallites size decreasing. Overall, the broadened

features of the Debye’s rings can be undoubtedly ascribed tothe presence of nanocrystallites with very small sizes.By increasing the incidence angle, the condition of diffraction

in reflection is restored and the corresponding ED pattern isreported in Figure 3c. From the analysis of the ED signals oneobtains the interplanar spacings (Table 1) that reveal thepresence of two concomitant sets of Debye’s rings, onecorresponding to nanocrystalline Si and the other onebelonging to the nanocrystalline NiSi2 phase having a cubicstructure with a lattice constant spacing a = 0.541 nm.22−24

Figure 3. Electron diffraction patterns (EDPs) in reflection (a, c) andtransmission (b) conditions from different regions of the sample: (a)boundary region between as-grown and plasma exposed areas:concomitant presence of spots (from the single crystal in the uncoatedregion) and Debye’s rings (from the randomly oriented Si crystallitesin the coated region); (b) outermost part of the plasma exposed area,constituted by protruding Si crystalline structures. In such conditions,the e-beam is practically parallel to the sample surface (assuming asreference that one of the as-grown regions) and interacts, intransmission, only with the protruding structures generated by theMW-RF plasma; (c) plasma exposed area: by tilting the incidenceangle, the EDP includes also the structural information coming from alarger depth, making possible the detection of a second crystallinephase due to the presence of Ni (rings with black dots) and identifiedas NiSi2, according to the data reported in Table 1. The indexing for Siand NiSi2 phase are reported as white and black labels, respectively.

Crystal Growth & Design Article

dx.doi.org/10.1021/cg3006212 | Cryst. Growth Des. 2012, 12, 4473−44784476

To better show the morphological and structural character-istics of the elongated deposits, some experiments have beencarried out under the same conditions, but using Si(100)substrates patterned by lithography with 150 nm Ni spots. Thedimension of the spots have been chosen in order to match thebasal size of the conical vertical nanostructures forming thearrays of Figure 1. A general view of a sample at the end of thedeposition process and details of a nanocone can be observedin Figure 4a,b, respectively.The individual conical objects shown in Figure 4a,b, more

than 1 μm long and with a base of approximately 100 nm,closely resemble the aligned vertical structures grown inside theNi-catalyzed areas of the “4” shaped samples (Figure 1).The results of the above-reported characterizations indicate

without any doubt that the conical forms generated on Sisubstrates using CH4 as feeding gas, under conditions that ingeneral produce one-dimensional carbon nanomaterials such asnanotubes and nanofibers, are instead made by polycrystallineSi. In the present case the only solid form of the C element isan amorphous hydrogenated layer covering the conical Sinanostructures, a layer that can be easily removed by a thermaltreatment.These unexpected and rather surprising results can be

explained considering that the growth processes are performedon Si substrates patterned with thin deposits of Ni, and that thesubstrates, under our experimental conditions, are negativelybiased. The highly negative value of the self-bias voltage,induced in the substrate holder by the application of an RFplasma, accelerates the motion of the positive H ions towardthe substrate. In a previous work dealing with the synthesis ofSi/CNT nanostructures, it was argued8 that the bombardmentby energetic H ions, combined with the substrate heating, couldinduce a process of mutual diffusion between Si and the Nideposits, giving rise to a molten Si/Ni phase rich in Si. Underthe action of the electrical field such a molten Si-rich phase iseasily pulled up along the vertical lines of the field, producingelongated Si-based entities. The presence of Ni is offundamental importance for the whole process, because theformation of Si/Ni droplets systems lowers the meltingtemperature of the bulk Si. The rapid cooling at the end ofthe process induces a crystallization of the melted Si inpolycrystalline structures. This is not only a speculation, basedon the fact that such elongated Si structures are formed only in

the presence of Si/Ni droplets produced by high biasing. TheRHEED results support the above-described liquid/solidmechanism of Si nanocones growth, occurring by Si extrusionfrom the substrate surface through the intermediate formationof melted Si/Ni compounds. The ED analysis highlights inparticular the following experimental findings:

(a) In the outermost part of the nanocones probed intransmission conditions (Figure 3b) only diffractionsignals belonging to Si phase are detected.

(b) In the intermediate layers (region in between from thetop part of the nanocones and the Si substrate) thepresence of small amounts of NiSi2 has been proven.

(c) No structural modification of the Si substrate has beenfound out of the region preliminarily Ni coated.

■ CONCLUSIONSWe demonstrated that conical Si structures can be realized onSi plates using a dual-mode MW/RF plasma reactor, in whichthe H ions experience an electric field biased toward the Sisubstrate, so that the ions may impinge on that surface withrather high kinetic energy.Precise location for the growth of the vertically protruding

structures can be achieved by a proper preparation of the Sisurfaces, modulating size and shape of the Ni layers sputteredinside patterned areas. Isolated nanocones with variable basaldiameters down to 100 nm or dense arrays formed by such

Table 1. Measured Interplanar Spacings in the RHEEDPattern shown in Figure 3 with the CorrespondingAttribution to Si (White Label in Figure 3) and NiSi2 (BlackLabel and • in Figure 3)a

attribution

measured interplanar distances d[nm] material d [nm] (hkl)

0.540 ± 0.005 Si 0.540 (100)0.315 ± 0.003 NiSi2 0.314 (111)0.311 ± 0.003 Si 0.312 (111)0.271 ± 0.003 Si 0.270 (200)0.243 ± 0.002 Si 0.242 (210)0.192 ± 0.002 Si 0.192 (220)0.164 ± 0.002 Si + NiSi2 0.163 (311) + 0.164 (311)0.125 ± 0.001 Si + NiSi2 0.124 (331) + 0.124 (331)

aThe attribution has been made on the basis of the PDF card #027-1402 and of the data reported in refs 20−22 for Si and NiSi2,respectively.

Figure 4. FE-SEM images of the sample at the end of the depositionprocess. (a) General view at low magnification; (b) magnification of asingle Si cone.

Crystal Growth & Design Article

dx.doi.org/10.1021/cg3006212 | Cryst. Growth Des. 2012, 12, 4473−44784477

elongated objects can be routinely prepared by engineering thedistribution of Ni sites on Si surfaces.The present approach allows one to fabricate ordered arrays

of Si nanocones strongly anchored to the Si substrates fromwhich the elongated structures have been extruded. Ourapproach appears more simple and effective in comparison withmore complicated methodologies involving two-step pro-cesses.25

The hypothesis for the building up of Si nanocones onto Ni-coated Si substrates is that such Si nanostructures originatefrom a molten/recrystallizazion process driven by the self-biasing of the substrate. The presence of Ni lowers the meltingtemperature of the crystalline Si substrate, due to the formationof a Si−Ni eutectic under both surface heating and activationvia intense ion bombardment.One collateral and surprising aspect of this work is that the

features of the Si cones and their assumed geometries strictlyremember images of similar objects reported in theliterature26−30 and commonly indicate as “carbon nanotubesor nanofibers”, in view of the fact that carbon-containing gaseswere used as reagents. References 26−30 are simply reported asexemplificative but not exhaustive cases. In general, such papersdid not report any structural analysis excluding the occurrenceof Si-containing rather than pure carbon nanostructure (or theirconcomitant presence). In such a context and on the basis ofthe present results, it is not possible to exclude that some of thealready reported C nanostructures using substrate bias PECVDtechniques could instead consist of (or containing) Si-basedcrystallites.Under the conditions of the present experiments a thin

surface carbon layer, produced by CH4 condensation, isproduced on the grown conical Si structures and on the Sisubstrate. The presence of amorphous carbon layers wasplanned to produce Si/C heterostructures with multifunctionalcapabilities. Thickness and structure of such C layers depositedon Si can be modulated by varying the composition of thefeeding gas phase and the duration of the synthesis runs.First of all, a thin C layer can protect the underlying Si

nanostructures from oxidation, and its complete removal iseasily achieved by a subsequent thermal treatment.Moreover, the hydrogenation of the carbon outermost

deposits may produce, in a controllable way, structural defects,and thus can be used to promote the formation of suitable sitesfor nucleation of other carbon forms, based on both sp2 and sp3

hybridization.31,32

Si/C systems hold considerable appeal as a way to improveperformances in electronic and optoelectronic applications, butthe realization of technologies based on Si architecturesdepends on the ability to control the growth and theassembling on an individual and collective basis.Overall, the proposed easily scalable one-step approach

appears suitable for the fabrication of complex 3D Siarchitectures or designed Si/C nanostructures to be integratedin planar Si devices.

■ AUTHOR INFORMATION

Corresponding Author*E-mail: marcorossi@uniroma1.it.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work has been partially supported by the Italian MIURthrough the PRIN2008 project: “Advanced nanomaterials andnanostructures for field- and photo-emission based devices”.

■ REFERENCES(1) Schmidt, V.; Wittermann, J. V.; Goesele, U. Chem. Rev. 2010,110, 361.(2) Bandaru, P. R.; Pichanusakorn, P. Semicond. Sci. Technol. 2010,25, 024003.(3) Zhu, J.; Yu, Z.; Burkhard, G. F.; Hsu, C.-M.; Connor, S. T.; Xu,Y.; Wang, Q.; McGehee, M.; Fan, S.; Cui, Y. Nano Lett. 2009, 9, 279.(4) C-Chen, P; Xu, J.; Chen, H.; Zhou, C. Nano Res. 2011, 4, 290.(5) Garnett, E.; Yang, P. Nano Lett. 2010, 10, 1082.(6) Riccitelli, R.; Di Carlo, A.; Fiori, A.; Orlanducci, S.; Terranova, M.L.; Santoni, A.; Fantoni, R.; Rufoloni, A. J. Appl. Phys. 2007, 102,054906.(7) Ng, S. H.; Wang, J.; Wexler, D.; Chew, S. Y.; Liu, H. K. J. Phys.Chem. C 2007, 111, 11131.(8) Toschi, F.; Orlanducci, S.; Guglielmotti, V.; Cianchetta, I.; Magni,C.; Terranova, M. L.; Pasquali, M.; Tamburri, E.; Matassa, R.; Rossi,M. Chem. Phys. Lett. 2012, 539−540, 94.(9) Martinu, J.; Klemberg-Sapieha, J. E.; Kuttel, O. M.; Raveh, A.;Wertheimer, M. R. Vac. Sci. Technol. A 1994, 12, 1360.(10) Koenig, H. R.; Maissel, L. I. IBM J. Res. Dev. 1970, 14, 168.(11) Tamor, M. A.; Vassel, W. C. J. Appl. Phys. 1994, 76 (6), 3823.(12) Tzakadze, Z. L.; Ostrikov, K.; Long, J. D.; Xu, S. Dia.mond Relat.Mat. 2004, 13, 1923.(13) Temple, P. A.; Hathaway, C. E. Phys. Rev. 1973, B7, 3685.(14) Cui, Y.; Lieber, C. Science 2001, 291, 851.(15) Bhattacharyya, S.; Samui, S. Appl. Phys. Lett. 2004, 84, 1564.(16) Gupta, R.; Xiong, Q.; Adu, C. K.; Kim, U. J.; Eklund, P. C. NanoLett. 2003, 3, 627.(17) Piscanec, S.; Cantoro, M.; Ferrari, A. C.; Zapien, J. A.; Lifshitz,Y.; Lee, S. T.; Hofmann, S.; Robertson, J. Phys. Rev. B 2003, 68,241312 R.(18) Adu, K. W.; Gutierrez, H. R.; Kim, U. J.; Eklund, P. C. Phys RevB 2006, 73, 155333.(19) Balkanski, M.; Wallis, F.; Haro, E. Phys. Rev. B 1983, 28, 1928.(20) Konstantinovic, M. J.; Bersier, S.; Wang, X.; Hayne, M.; Lievens,P.; Silverans, R. E.; Moshchalkov, V. V. Phys. Rev. B 2002, 66, 161311.(21) Rossi, M.; Vitali, G.; Karpuzov, D.; Kalitzova, M.; Budinov, H. J.of Mat. Science 1991, 26, 3337.(22) Pearson, W. B. Handbook of Lattice Spacing and Structures ofMetals and Alloys; Pergamon: New York, 1958.(23) Pearson, W. B. The Crystal Chemistry and Physics of Metals andAlloys; Wiley: New York, 1972.(24) Franciosi, A.; Weaver, J. H.; Schmidt, F. A. Phys. Rev. B 1982,26, 546.(25) Medvid, A.; Onufrijevs, P.; Mychko, A. Nanoscale Res. Lett.2011, 6, 582.(26) Semet, V.; Vu Thien Binh, P.; Vincent; Guillot, D.; Teo, K. B.K.; Chhowalla, M.; Amaratunga, G.A. J.; Milne, W. I.; Legagneux, P.;Pribat, D. Appl. Phys. Lett. 2002, 81, 343.(27) Kato, T; Jeong, G.-H; Hirata, T; Hatakeyama, R Thin Solid Films2004, 457, 2−6.(28) Tsakadze, Z. L.; Levchenko, I.; Ostrikov, K.; Xu, S. Carbon2007, 45, 2022−2030.(29) Denysenko, I. B.; Xu, S.; Long, J. D.; Rutkevych, P. P.;Azarenkov, N. A.; Ostrikov, K. J. Appl. Phys. 2004, 95, 2713.(30) Han, J.-H.; Choi, S. H.; Lee, T. Y.; Yoo, J.-B.; Park, C.-Y.; Kim,H. J.; Han, I.-T.; Yu, S. G.; Yi, W.; Park, G. S.; Yang, M.; Lee, N. S.;Kim, J. M. Thin Solid Films 2002, 409, 126.(31) Barnard, A. S.; Terranova, M. L.; Rossi, M. Chem. Mater. 2005,17, 527.(32) Terranova, M. L.; Orlanducci, S.; Fiori, A.; Tamburri, E.; Sessa,V.; Rossi, M.; Barnard, A. S. Chem. Mater. 2005, 17, 3214.

Crystal Growth & Design Article

dx.doi.org/10.1021/cg3006212 | Cryst. Growth Des. 2012, 12, 4473−44784478